Edwards' concept of the alcohol dependence syndrome (ADS) has served as the stimulus for much alcohol research and has provided the basis for the definitions of alcohol dependence in DSM-III-R and the upcoming ICD-10. Epidemiologic research based on the ADS has begun, and more can be expected. However, very little is known about the reliability, validity, or prognostic implications of the new definitions of alcohol dependence in general population samples, or in women from any setting. The proposed study is designed to provide information on these issues. Sixteen hundred household residents with a recent occasion of at least moderately heavy drinking will be recruited. They will be evaluated further for alcohol consumption, alcohol dependence and other alcohol-related problems with an assessment procedure suitable for epidemiologic research, administered by non-clinician interviewers. The male/female ratio in the sample will be adjusted to provide approximately equal numbers of males and females with current or recent alcohol dependence. A subset of the sample will be re- interviewed after one year, to determine follow-up status on many of the variables initially assessed. Short-term aims include investigation of the reliability of DSM-III-R and ICD-10 definitions of alcohol dependence and concordance with earlier definitions of alcoholism or alcohol dependence. More important short-term aims include investigation of the validity of alcohol dependence. This will be accomplished in three components: (1) analyses of the association of alcohol dependence diagnoses with a set of predicted correlates (external to symptoms of the diagnoses) constituting validating or criterion variables, (2) comparison of fully-structured assessment procedures for alcohol dependence with clinician evaluations, and (3) evaluation of the short-term (one-year) predictive validity of alcohol dependence diagnoses by examining chronicity/remission in conjunction with drinking status and environmental variables. Analyses of follow-up status will determine if alcohol dependence as defined conforms in a general population sample to predictions made from theory about its course and natural history. Long-term aims include following the sample over several years, to determine predictors of onset of alcohol dependence in those at risk who have not developed the condition, and also to further investigate remission/chronicity.